1.5. Objetivos de la investigación
2.2.8. Estructura y tipos de texto
Retinal breaks can be closed by two methods; by scleral buckling, which opposes the break against the RPE, and by internal tamponade, with agents such as air, gases and silicone oil. These techniques are often combined, particularly in complicated RDs. In vitrectomised eyes scleral buckling alone does not usually result in closure of retinal breaks and internal tamponade is also needed. It is thought that in non-vitrectomised eyes the vitreous face may play a part in break closure after scleral buckling. However, this seems an incomplete explanation as, after buckling "U" tears, the flap of the tear, which is the site of vitreous attachment, often remains elevated implying that the vitreous is not opposed against the retina(71).
All the agents that are used routinely for internal tamponade have lower specific gravities than intraocular fluids and are therefore buoyant. With accurate posturing of the patient after surgery the gas or air bubble (or silicone oil) prevents recruitment of SRF and flattens the retinal break against the RPE. Breaks in superior retina are more readily closed than those in inferior retina. Retinal breaks can also be closed by tamponading agents even if they are not flattened against the RPE because of surface tension effects of the tamponading agents, although this is not the predominant mechanism(75).
Temporary tamponade is achieved using air or gases, and prolonged or permanent tamponade with silicone oil. Intravitreal air was advocated by Rosengren in 1938(76) (from Chawla 1973) and popularised in this country by Chawla(76). Air is rapidly absorbed from the posterior segment and if break sealing has not occurred by the time the bubble disappears, the break can reopen and the detachment recur. A longer acting gas, sulphur hexafluoride (SF6) was introduced in 1973 to avoid this problem(77). SF6 absorbs nitrogen and oxygen from the blood, expands
if injected in a pure form and reabsorbs more slowly than air. More recently the perfluorocarbon gases (perfluoroethane, C2F6, and perfluoropropane, C3F8) have been introduced (78,79). These gases persist longer than SF6 gas(80) and tamponade can be pro-longed by repeat injections. The advantage of the longer acting gases is that the extended period of break closure gives time for a chorioretinal scar to form, which seals the break.
Cataract is the most frequent complication of internal tamponade with SF6 and C3F8 gas. Lens opacities are more likely to develop if the gas remains in prolonged contact with the lens, are usually posterior subcapsular and reversible, but can be permanent(79).
Raised intraocular pressure can develop during the post-operative period if too large a volume of undiluted gas is injected or, in aphakic eyes, if peripheral anterior synechiae should form. Air travel is contraindicated as the bubble of gas will expand in the reduced atmospheric pressure of the cabin(81).
Gas may pass through retinal breaks, becoming trapped in the subretinal space. In non-vitrectomised eyes the bubble can pass into the retrohyaloid space, increasing the traction exerted on the retinal break, or producing new tears(82).
The use of silicone oil in RD surgery was advocated by Cibis and popularised by Scott(83,84). Cibis described its use in 33 patients; 5 had giant retinal tears, 13 PVR and the remainder had multiple breaks and retinal atrophy. Partial or complete retinal reattachment was achieved in 23 (70%) of these otherwise inoperable cases(83). Silicone oil is currently used for selected cases, such as those with advanced PVR or static vitreo-retinal traction with posterior breaks, most giant retinal tears, selected inferior breaks, cases with intractable breaks and to prevent severe hypotony. Retinal reattachment takes place because silicone oil closes retinal breaks and not because the retina is forced back against the RPE(85). This clinical impression is supported by experimental studies as silicone oil did not prevent the development of purely tractional RDs in animal models of PVR(86,87). In combined traction/ rhegmatogenous RDs, where it may not possible to relieve V- R traction completely, silicone oil can be used to treat the rhegmatogenous component of the detachment, leaving an area of purely traction RD. This is known as "rhegmatogenous confinement" (Figure 1.3).
Initially silicone oil was injected into the mid-vitreous cavity or retrohyaloid space. The vitreous gel become compressed against the back of the lens. Closed microsurgical techniques were introduced in the 1970s(88), and since then silicone
\ J /
B
Figure 1.3 Diagramatic representation of "rhegmatogenous confinement".
A Vitreoretinal traction (indicated by arrows) and a retinal break causing combined traction/rhegmatogenous retinal detachment B "Rhegmatogenous confinement" after treating the retinal break with a patch
oil has been used after vitrectomy. This means that the oil is in close contact with the lens and posterior segment tissues. Silicone oil has a high surface tension and in the eye tends to form a large, single bubble. In aphakic eyes the bubble has the potential to pass into the anterior chamber of the eye by occluding aqueous flow through the pupil (silicone oil pupil block).
Complications were noted soon after silicone oil was introduced. In 1963 Shea reported glaucoma, keratopathy and lens changes, and cataract and glaucoma were observed by Dufour in 1968(89)(from Cockerham 1970). In 1967 Watzke noted similar complications, and cautioned against the use of silicone oil(90). The pathogenesis of complex RD and some of the reasons why silicone oil induces complications are now better understood. These factors, combined with advances in surgical techniques and careful case selection have resulted in a lower incidence of silicone oil-related complications. However, potentially sight threatening complications still occur (Table 1.1)(89 -105). This has led to reluctance on the part of many surgeons to use silicone oil, particularly in the USA(106).
The varying incidence of complications shown in Table 1.1 may, in part, be explained by case selection. The study by Antoszyk, for instance, which has a high incidence of keratopathy, was a study of trauma cases, where there may have been pre-existing corneal damage, or zonular rupture which allowed silicone oil to pass into the anterior chamber(IOI).
Silicone oil keratopathy, which is more common in aphakic eyes, is characterised by corneal decompensation and band keratopathy. It develops in eyes where silicone oil is or has been in contact with the corneal endothelium. In an animal model endothelial cell loss was observed in eyes where silicone oil was in contact with the cornea. This was thought to have occurred because the oil provided a barrier to nutrition(107).
Lens opacities eventually develop in the majority of silicone oil filled phakic eyes. Histological study has shown silicone oil laden macrophages on the lens capsule, but no silicone oil in the substance of the lens(92). Lens opacities are also thought to occur because silicone oil provides a mechanical barrier to nutrition.
Several different forms of glaucoma can occur in silicone oil filled eyes. Pupil block glaucoma may develop in aphakic eyes if the anterior face of the oil bubble blocks the peripheral iridectomy and pupil. Chronic angle closure glaucoma can occur secondary to the formation of peripheral anterior synechiae, and open angle
glaucoma may be aggravated by the presence of silicone oil laden macrophages in the trabecular meshwork(92).
There are other, more contentious issues regarding the use of silicone oil, particularly in relation to whether silicone oil stimulates reproliferation of ERMs and whether it is toxic to the retina. Cockerham and Watzke both noted the presence of episilicone membranes, but because ERMs were not surgically removed at that time little conclusion could be drawn(89,90). Lewis found reproliferation of ERMs in 19 of 31 eyes (61%), 15 of which subsequently developed recurrent RD(108). Riedel noted reproliferation of epiretinal and subretinal membranes in 40% of his series of 415 patients(103), Federman reported an incidence of 15%(100) and other authors between 3 and 71%(100)(from Federman 1988). The clinical evidence suggests that silicone oil is associated with reproliferation of ERMs, but whether the relationship is causal has yet to be clarified.
This issue has been addressed using animal models of PVR and tissue culture experiments. Cell injection techniques provide a good model for investigating different aspects of PVR. Fastenberg and Lean(109,110) found no difference in the rate of ERM formation in vitrectomised rabbits' eyes with and without prior silicone oil injection. These findings are at variance however, with those of Lambrou, but he used a different experimental d e s ig n (lll). Tissue culture experiments have been undertaken to see whether silicone oil has any effect on cell types implicated in the formation of ERMs. Fibroblasts remain viable when plated directly onto silicone oil (112) and human RPE cells grown under silicone oil dedifferentiate and proliferate (113). These studies provide evidence that silicone oil may encourage the development of ERMs by a direct effect on cell behaviour. Lambrou has postulated that silicone oil may stimulate the retina or residual ERM tissue to produce growth fa c to rs (lll). On the other hand the effect of silicone oil may be purely mechanical. Many factors and processes have been implicated in the pathogenesis of PVR, e.g. inflammation, RPE cell proliferation and effects of growth factors and FN. A bubble of silicone oil will tend to localise these factors in the preretinal space, encouraging ERM formation.
The question of silicone oil retinotoxicity has provoked much debate and stimulated a considerable volume of research, often with conflicting results. Retinae from enucleated silicone oil filled eyes have marked degenerative changes. Flowever, firm conclusions cannot be drawn from this as the eyes were usually surgical failures. Clinical evidence is difficult to interpret as it is hard to distinguish changes due to preexisting retinal pathology from those which could be attributable to toxic effects of silicone oil.
Table 1.1 Complications of internal tamponade with silicone oil
Date Author Ref N FU K'top Glauc. Cat. Other
% % % % 1967 Watzke 90 51 ND ND 15 89 72 RD 1970 Cockerham 89 78 1 mn 11.5 ND 35 91 RD 1973 Kanski 91 36 > 6 mn 9 5.5 33 50 RD 1979 Leaver 92 93 12 mn 6.5 14 65 46 RD 1980 Haut 93 200 8 mn 18.5 12.5 50 ND 1985 McCuen 94 164 6 mn 29 10 67 42 RD 37 hyp 1986 Cox 95 51 6 mn 22 12 50 35 RD 1986 Chan 96 407 1 mn- 23yrs 12 17 62 33 RD 1987 Laqua 97 500 6 mn 23 a 1p 1 7 a 4 p 100 23 RD 1987 Ando 98 101 36 mn 34 a 7 p 16 100 ND 1987 Yeo 99 52 ND 25 6 39 30 RD 1988 Federman 100 150 6 mn 15 10 100 22 1989 Antoszyk 101 42 6 mn 48 0 ND 50 RD 50 hyp 1989 Laganowski 102 44 Bl 44 SI at 6 mn at 6 mn 14 39 29 43 ND ND ND 1989 Riedel 103 415 6 mn 6 6 100 40 PVR 1992 SSG 1 104 52 24 mn 21 0 ND 20 RD 11 hyp 1992 SSG II 105 Gp 1 64 18 mn 30 1.5 ND 36 RD 20 hyp 105 Gp 1 63 18 mn 40 3 ND 39 RD 22 hyp
N = number of patients a = aphakic FU = minimum follow up p = phakic Glauc = glaucoma Cat = cataract K'top = keratopathy hyp = hypotony RD = persistent retinal detachment mn = months Bl = basal peripheral iridectomy ND = no data SI = superior iridectomy yrs = years
Early studies using electrodiagnostic tests (electro-retinogram, ERG and electro oculogram, EOG) complicated the issue. Silicone oil filled eyes were found to have almost absent ERGs and EOGs. This was interpreted as evidence of retinal toxicity(114)(from Foerster 1985). However, the findings are probably explained by the fact that silicone oil is an electrical insulator(92). Foerster's clinical studies showed "no consistent changes that could be attributable to a delayed retinotoxic effect" (114) and vision-evoked cortical potentials undertaken on patients before and after removal of silicone oil showed no evidence of optic nerve damage(115).
Experimental studies using animal models have produced conflicting results. In a study by Meredith, ERGs were performed before and after silicone oil or BSS were injected into vitrectomised rabbit eyes. After an initial postoperative decline in B wave amplitude in both groups of animals postoperative values reached preoperative values in both groups by 6 months. Histological study of the retinas from these animals showed little abnormality in either group(116). Similar findings were reported by Ober (117), but are in sharp contrast to the findings of other studies. The rabbit retinae in a study by Gonver showed pronounced degenerative changes, particularly in the outer plexiform layer. The changes were more obvious in the upper retina, where the silicone oil bubble was in direct contact with the retina. Changes were also present when purified silicone oil was used(118). In a similar study on monkeys, consistent degenerative changes were seen in all layers of the retina, particularly in the periphery. These changes were not detected in control eyes which had had injection of BSS rather than silicone oil(119). In another study using primates, light and electron microscopy showed only minimal changes, which were the same in silicone filled eyes and controls and which were attributed to surgical trauma(120). There are several possible explanations for the different findings in these studies. Firstly, different species were used, secondly, surgical techniques were not the same in each study, and finally, silicone oils of varying chemical composition were used.
Although the evidence for retinotoxic effect of silicone oil is controversial there are several theoretical reasons why silicone oil may produce histological changes or functional abnormalities.
Silicone oil is a polydimethylsiloxane, made by polymerisation of dimethylsiloxane- oligomeres. The viscosity is determined by the polymer chain length i.e. the molecular weight, but a statistical distribution of molecular weights is an invariable result of the polymerisation process. Viscosity can also be altered by mixing oils of differing molecular weights. The silicone oil that is used clinically is 'medical grade', and is manufactured by several different companies. Physicochemical analysis has
been undertaken of oils from different sources, to determine the molecular weight, viscosity, the percentage of low molecular weight components and the percentage of volatile components (low molecular cyclosiloxanes which vaporise at 37°C).
Silicone oils with viscosities of 1,000 and 5,000 mPas have been analysed (121,122). These studies showed low molecular weight components and volatile cyclosiloxanes in varying amounts. The significance of these findings are that low molecular weight components are thought to lower oil émulsification thresholds. The very low molecular weight, volatile compounds are small enough to diffuse into tissues, and may be responsible for some of the complications observed.
In theory, lipophilic substances will dissolve in silicone oil and this has been investigated by Refojo(123). In an experimental animal model silicone oil was found to extract retinol, with concentrations increasing with time. He also analysed silicone oil removed from patients' eyes and found both cholesterol and retinol to be present. It is not known what effect lipid extraction may have on ocular tissues in terms of structure and function.
Silicone oil has also been analysed for the presence of impurities. Parel analysed 18 different oils and found impurities in amounts varying from 350-4,900 ppm. Arsenic, antimony and cyanide were found amongst others (Parel, unpublished data, presented at ARVO 1989).
Other complications associated with silicone oil include émulsification and refractive errors. Emulsification is seen in most cases, varying from a few tiny droplets in the anterior chamber to an 'inverted hypopyon' or an anterior chamber completely filled with milky, emulsified oil. Emulsification occurs when two immiscible fluids are agitated in the presence of a third substance (an emulsifier, such as proteins). Intraocular proteins may activate émulsification, encouraged by eye movements(75).
Refractive changes occur because silicone has a higher refractive index than intraocular fluids; the + or - change being principally determined by the shape of the anterior surface of the bubble(93,124).